The objective is to identify patient characteristics that can be used to predict successful response to behavioral treatments for the long term control of obesity. A research design and procedures are developed that address methodological problems in previous research that have interfered with the accumulation of replicable findings concerning this issue. Two hundred consecutive patients in a behavioral treatment program for weight control are used as subjects. Data are collected from these subjects prior to entry in a one year treatment program, at numerous points during the course of the program, at program termination, as well as at 6, 12 and 18 month post program followups. Data collected operationalizes the major factors identified in the literature. Emphasis is placed on instrumented, behavioral and clinically practical modes of data collection. Hypotheses not previously addressed in the literature are developed and tested. A mathematical modeling technique is used to develop and test more complex hypotheses concerning interactions among variables and sequences of causal events in which some variables affect outcomes through the effect of mediating variables. This modeling technique is proposed as a potentially useful method to evaluate individual differences in response to treatment for other types of therapeutic interventions such a treatment for substance abuse. A new measure that can be used to preduct treatment outcome in behavioral treatment for obesity is developed.